Brownfields and Environmental Justice: Income,Education,and Race

Introduction

The United States Environmental Protection Agency (USEPA) defines environmental justice as “everyone <enjoying> the same degree of protection from environmental and health hazards and equal access to the decision-making process to have a healthy environment in which to live,” regardless of race or income. ¹ According to the USEPA, many brownfields contain pollutants that are hazardous to human and environmental health. ² Unlike environmental hazards associated with currently operating facilities, such as power plants and manufacturing industries, brownfields afford unique opportunities to correct environmental injustices by removing environmental or health risks. There are at least 450,000 brownfields in the United States. ³ Because these sites are inactive, they can be cleaned and restored to an active use that benefits the community.

If a goal of brownfield redevelopment is to right an environmental injustice, then an understanding of the populations that are disproportionately exposed to brownfields is necessary. To determine demographics around brownfields we examined the sites catalogued in the Assessment, Cleanup and Redevelopment Exchange System (ACRES) database maintained by the USEPA, which contains information on grants awarded under programs administered by this agency. ⁴ The criteria used in selecting the sites for funding suggest that the database is not representative of all brownfields in the country, but is biased towards sites with demonstrated community need and the potential to stimulate the local economy or create public greenspace. ⁵

Using the January 2008 ACRES database, we selected only sites funded in 2003 or later. Those sites were subject to the Small Business Liability Relief and Brownfields Revitalization Act of 2002, which requires applicants to provide information about the site's location, history, potential contamination, and future plans. ⁶ We removed duplicate entries marking different stages in redevelopment, sites with obvious errors in latitude and longitude, and sites in U.S. territory outside the fifty states and the District of Columbia, leaving 3,447 sites for analysis. The distribution of sites around the continental United States is shown in Figure 1.

OPEN IN VIEWER

FIG. 1.

Sites located in the contiguous 48 states used in the analysis of socioeconomic conditions around EPA-funded brownfields. These 3,447 sites were reduced to 1,343 census tracts with brownfield sites in the 2000 census. Many brownfields are clustered in urban areas so that individual points on the map are not always visible.

Discussion

We used census data at the tract level to analyze the socioeconomics and demographics around brownfield sites. Census tracts offer high resolution without compromising data availability. ⁷ Using geographic information system (GIS) software, we found the census tract hosting each brownfield site and associated those demographics with the site. This method, the unit-hazard coincidence method, sees common use in environmental justice research, wherein census tracts are considered an acceptable unit of analysis. ⁸ We chose this method rather than defining a radius of influence around each brownfield, because we had no basis for estimating a radius of influence of a brownfield site. When analyzing demographics, we counted each tract with at least one brownfield site only once regardless of the number of sites in the tract. Under this procedure, contiguous sites and sites close enough to have likely similar sources and effects on the surrounding community are not overrepresented in the analysis. This method did not compromise resolution, as 940 of the 1,343 tracts with brownfields had only one site in them, and 1,149 tracts contained fewer than three sites (2000 Census).

We chose four categories of indicators to analyze: race, age distributions (suggesting workforce size), education levels, and economic conditions. Most of the indicators were prepared by normalizing census data against the appropriate population. Any unavailable or undefined indicators for a given census tract were zeroed. We used a Z-test to evaluate significance by comparing the census tracts with brownfields to the entire population of census tracts. The requirements of the Z-test are met in this application because the population standard deviations of the indicators are known. Though the distributions of the indicators themselves are not normal, the sample size is sufficiently large that the distribution of the Z statistic will be approximately normal under the null hypothesis. Making no assertion about the direction of the difference, we tested the null hypotheses, H₀, that there was no difference between the mean value of the indicator in census tracts with brownfields and all census tracts, with the alternative, H_a, that the means were different. Table 1 presents the results for the Census year 2000. We also analyzed median household income. The mean (±standard deviation) of this indicator in census tracts with EPA-funded brownfields is $34,100 ± $14,400 compared with $44,200 ± $20,800 in all census tracts. Performing the same Z-test on this indicator yields a p-value of essentially zero, suggesting that median household incomes are significantly lower in tracts with brownfields.

Based on a 95% confidence level, p-values less than 0.05 suggest that there are significant differences in the indicator levels between census tracts with USEPA-funded brownfields and all of the census tracts in the nation. Several indicators suggest that the areas around brownfields are economically depressed compared with the entire United States including the poverty rate, the unemployment rate, the proportion of vacant housing units, and the median household income. While significant differences exist between the distributions of race around brownfields and throughout the entire nation, the economic and educational differences are even more pronounced. Furthermore, people living around brownfields are significantly less educated at both high school and college levels than they are in the average census tract.

Conclusions

Brownfield are generally not in active use. Therefore, redevelopment offers a unique opportunity to reduce environmental and health risk without disturbing a community's social or economic dependence on the site. According to USEPA, the most important method of addressing environmental injustice in environmental policy, including site-specific brownfield remediation strategies, is community involvement. ⁹ One example of a successful, community-based restoration effort is the removal of abandoned mine drainage water along Herron Avenue in Pittsburgh, Pennsylvania. Waters from an abandoned mine under the area repeatedly flooded a church in this underprivileged section of the city. Cooperation between the congregation, community members, and state agencies facilitated the installation of a drainage system that prevents flooding in the church and reduces the risk of human contact with the effluent. Furthermore, the church was able to take advantage of the constant temperature of the abandoned mine drainage water and use it to support a geothermal heating and cooling system which cuts heating costs for the church by seventy-five percent. ¹⁰

An examination of the details surrounding the remediation efforts in the Herron Avenue Corridor reveals that community involvement was instrumental in developing a remediation plan that did more than merely address the environmental hazard. This project also demonstrates the successful achievement of two goals that should be part of brownfield remediation plans. First, the restoration efforts should reduce the contamination at the site. Second, restoration should attempt to bring some of the well-understood economic benefits of brownfield remediation to the community in a way that improves the lives of residents. ¹¹ It is in achieving the second goal that community involvement is most critical. Without input from community groups, local businesses, and local governments at all stages in the redevelopment effort, developers cannot know how proposed remediation plans will affect residents.